Previous work in literature has indicated that template of web pages represent noisy information in web collections, and advocate that the simple removal of template result in improvements in the quality of results provided by Web search systems. In this paper, we study the impact of template removal in two distinct scenarios: large scale web search collections, which consist of several distinct websites, and intrasite web collections involving searches inside of web sites. Our work is the ï¬rst in literature to study the impact of template removal to search systems in large scale Web collections. The study was carried out using an automatic template detection method previously proposed. As contributions, we present statistics about the application of this automatic template detection method to the well known GOV2 reference collection, a large scale Web collection. We also present experiments comparing the amount of template detected by our automatic method to the ones obtained when humans select templates. And ï¬nally, experiments which indicate that, in both experimented scenarios, template removal does not improve the quality of results provided by search systems, but can play the role of an effective loss compression method by reducing the size of their indexes.
